Novel N-(4,5,6,7-tetrahydrobenzisoxazol-4-yl)amides as HSP90 inhibitors: design, synthesis and biological evaluation

Abstract

Novel N-(4,5,6,7-tetrahydrobenzisoxazol-4-yl)amide derivatives were designed and synthesized as potential HSP90 inhibitors. The synthetic pathway commenced with 6,7-dihydrobenzo[d]isoxazol-4(5H)-ones, utilizing the Ritter reaction as a key step. Molecular docking, molecular dynamics simulations, and MM/GBSA analysis guided the selection of compounds for synthesis and provided insights into the interaction mode of the most active compound with HSP90α. The synthesized compounds exhibited significant antiproliferative effects against breast cancer cell lines ERα+ MCF7 and HER2+ HCC1954. Lead compounds with submicromolar IC₅₀ values, initially synthesized as racemates, were subsequently obtained and tested in their enantiopure forms. In HER2+ HCC1954 cancer cells, the molecular pathways regulated by compound (R)-8n were characterized. Treatment with compound (R)-8n resulted in the pronounced suppression of HSP90-related pathways, including key oncoreceptors (HER2, EGFR, c-MET) and mitogenic kinases (AKT, CDK4). Additionally, compound (R)-8n induced apoptosis, as evidenced by the accumulation of cleaved PARP. The inhibitory effect of compound (R)-8n on the HSP90 pathway was corroborated by molecular modeling and further validated through the observed suppression of client proteins, along with an upregulation of HSP70, a well-established marker of HSP90 inhibition. The activity of compound (R)-8n was associated with cell cycle arrest at the G2/M phases, ultimately leading to dose-dependent cell death. Notably, compound (R)-8n demonstrated substantial selectivity toward breast tumor cells. These findings suggest that N-(4,5,6,7-tetrahydrobenzisoxazol-4-yl)amides represent a promising class of HSP90 inhibitors for anticancer therapy.

Fig. 1. The conformations of the NTD lid of HSP90 in the complexes with 17-DMAG (pdbid 1OSF) (A), luminespib (pdbid 6LTI) (B), pimistepib analog (pdbid 5ZR3) (C), KUNA-111 (pdbid 7UR3) (D) and structures of these and related inhibitors.

Chart 1. Structures of HSP90-targeting compounds containing benzisoxazole or partially saturated benzisoxazole core.

Chart 2. Structures of 6,7-dihydrobenz[d]isoxazol-4(5H)-ones 1a and 6,7-dihydrobenz[c]isoxazol-4(5H)-ones 1b – perspective scaffolds for the preparation of new HSP90 inhibitors.

Fig. 2. Structures of new potential classes of HSP90 inhibitors (R)-athbzi (A) and (S)-athbzi (B) and superposition (C) of theoretical poses of (R)-athbzi-2 (green) and (S)-athbzi-2 (pink) with experimental pose of luminespib (teal) in ATP-pocket of the NTD HSP90.

Scheme 1. Synthesis of potential HSP90 inhibitors 7–9.

Scheme 2. Resolution of 3-(5-isopropyl-2,4-dimethoxyphenyl)-4,5,6,7-tetrahydrobenzo[d]isoxazol-4-amine 4a.

Fig. 3. Signaling pathways in HCC1954 breast cancer cells treated with (R)-8n. GAPDH antibodies were used as a loading control. LUMI – luminespib (NVP-AUY922) (A). HER2 (ERBB2) and EGFR as key HSP90 clients (diagram of HSP90 signaling pathways created using STRING) (B).

Fig. 4. Cell cycle distribution of HCC1954 cells treated by compound (R)-8n at concentrations of 0 (control, DMSO), 0.125, 0.5 and 2 μM for 48 h.

Fig. 5. The time dependence of the RMSD of (R)-8n. Non-hydrogen atoms of (R)-8n and the backbone atoms of M^pro were used in the calculation (A); the change of MM/GBSA ΔH value during the MD simulation of the predicted complex of (R)-8n with HSP90α (B); 3D representation of the predicted binding pose of (R)-8n in complex with N-terminal of HSP90α (hydrogen bonds shown as yellow dash lines) (C); 2D representation of intramolecular interactions of (R)-8n with N-terminal of HSP90α (D).

Fig. 6. Antiproliferative activity and toxicity of compound (R)-8n. The breast cancer cells and fibroblasts were treated with the compound and after 72 hours their survival was evaluated by MTT test.